[ovs-dev] [PATCH] OVN resource agent - make promotion synchronous

Currently inside the ovsdb_server_promote() function we call 'promote_ovnnb' and 'promote_ovnsb' and then just record the new master state in the CIB. This creates a race because those two promote commands are asynchronous so when we exit the ovsdb_server_promote() function the underlying DBs are

Re: [ovs-dev] [PATCH v3 2/5] dpif-netdev: Trigger parallel pmd reloads.

Hello Ilya, On Mon, Jul 8, 2019 at 4:55 PM Ilya Maximets wrote: > On 04.07.2019 14:59, David Marchand wrote: > > pmd reloads are currently serialised in each steps calling > > reload_affected_pmds. > > Any pmd processing packets, waiting on a mutex etc... will make other > > pmd threads wait for

Re: [ovs-dev] [PATCH] OVN resource agent - make promotion synchronous

Thanks a lot Michele. Just mentioning that this has been tested in an OpenStack environment successfully. A timeout is not needed for the while loop since pacemaker will enforce its own. On Tue, Jul 9, 2019 at 9:20 AM Michele Baldessari wrote: > > Currently inside the ovsdb_server_promote() funct

[ovs-dev] Вам ведь это интересно?

Просим прощения за беспокойство, думаю, мы сможем быть для Вас полезны. Мы, экспертная команда веб-разработчиков по CMS WordPress (эта система управления занимает 34% среди всех сайтов мира). Владеем самыми крупными сообществами по данной тематике в России и СНГ. Создаем сайты любой сложности,

Re: [ovs-dev] [PATCH v3 2/5] dpif-netdev: Trigger parallel pmd reloads.

On 09.07.2019 10:14, David Marchand wrote: > Hello Ilya, > > On Mon, Jul 8, 2019 at 4:55 PM Ilya Maximets > wrote: > > On 04.07.2019 14:59, David Marchand wrote: > > pmd reloads are currently serialised in each steps calling > > reload_affected_pmds. >

Re: [ovs-dev] [PATCH] OVN resource agent - make promotion synchronous

On Tue, Jul 9, 2019 at 1:04 PM Daniel Alvarez Sanchez wrote: > Thanks a lot Michele. > Just mentioning that this has been tested in an OpenStack environment > successfully. A timeout is not needed for the while loop since > pacemaker will enforce its own. > > On Tue, Jul 9, 2019 at 9:20 AM Michel

Re: [ovs-dev] [PATCH] OVN resource agent - make promotion synchronous

Thanks for the patch Michele. As @Daniel pointed out, this have been tested and works. On Tue, Jul 9, 2019 at 8:18 AM Michele Baldessari wrote: > > Currently inside the ovsdb_server_promote() function we call 'promote_ovnnb' > and 'promote_ovnsb' and then just record the new master state in the

Re: [ovs-dev] [PATCH] OVN resource agent - make promotion synchronous

Bleep bloop. Greetings Michele Baldessari, I am a robot and I have tried out your patch. Thanks for your contribution. I encountered some error that I wasn't expecting. See the details below. checkpatch: WARNING: Line is 89 characters long (recommended limit is 79) #50 FILE: ovn/utilities/ovn

Re: [ovs-dev] [PATCH v3 2/5] dpif-netdev: Trigger parallel pmd reloads.

On Tue, Jul 9, 2019 at 9:36 AM Ilya Maximets wrote: > On 09.07.2019 10:14, David Marchand wrote: > > We can skip the non-PMD thread in this loop for consistency with > dp_netdev_reload_pmd__() if this is what you meant. > > > > But for a non-PMD thread, pmd->reload is set to false at init and is

Re: [ovs-dev] [PATCH v2] Avoid dp_hash recirculation for balance-tcp bond selection mode

> > Thanks for the patch. > > I wasn't able to apply this patch. In two places, the patch was corrupted > (the > patch didn't properly represent page-break characters in the source code but > transformed them into blank lines). I fixed that up after figuring out what > was > going on, but aft

[ovs-dev] [PATCH v3] Balance-tcp bond mode optimization

v1->v2: Updated datapath action to hash + lb-output. Updated throughput test observations. Rebased to OVS master. v2->v3: Rebased to OVS master. Fixed git merge issue. Vishal Deep Ajmera (1): Avoid dp_hash recirculation for balance-tcp bond selection mode datapath/linux/compat/include/li

[ovs-dev] [PATCH v3] Avoid dp_hash recirculation for balance-tcp bond selection mode

Problem: In OVS-DPDK, flows with output over a bond interface of type “balance-tcp” (using a hash on TCP/UDP 5-tuple) get translated by the ofproto layer into "HASH" and "RECIRC" datapath actions. After recirculation, the packet is forwarded to the bond member port based on 8-bits of the d

Re: [ovs-dev] [PATCH] ofproto: Preserve ofport number for failed ofport across restarts

> > I've been taking a look at this patch for the last few minutes. It > introduces a lot > of mechanism for the use case. Did you consider any simpler mechanisms to > achieve the same effect? What prevented them from working? > I agree Ben. This change does bring some complexity to the code.

Re: [ovs-dev] [PATCH v3] Avoid dp_hash recirculation for balance-tcp bond selection mode

On 09.07.2019 20:02, Vishal Deep Ajmera wrote: > Problem: > > In OVS-DPDK, flows with output over a bond interface of type “balance-tcp” > (using a hash on TCP/UDP 5-tuple) get translated by the ofproto layer into > "HASH" and "RECIRC" datapath actions. After recirculation, the packet is

Re: [ovs-dev] [PATCHv14 2/2] netdev-afxdp: add new netdev type for AF_XDP.

Hi William, I tried running my PVP test, and with this patchset, I could not see the deadlock I used to see with all the previous patches. However there is still the issue that the maximum packet size supported is 1028 bytes, all larger packets are dropped!! Unfortunately, I’m going on PTO

Re: [ovs-dev] [PATCH v3] Avoid dp_hash recirculation for balance-tcp bond selection mode

Bleep bloop. Greetings Vishal Deep Ajmera, I am a robot and I have tried out your patch. Thanks for your contribution. I encountered some error that I wasn't expecting. See the details below. build: libtool: compile: gcc -std=gnu99 -DHAVE_CONFIG_H -I. -I ./include -I ./include -I ./lib -I .

[ovs-dev] [PATCH] ovsdb-server: drop all connections on read/write status change

Prior to this patch, only db change aware connections were dropped on a read/write status change. However, current schema in OVN does not allow clients to monitor whether a particular DB changes this status. In order to accomplish this, we'd need to change the schema and adapting ovsdb-server and e

Re: [ovs-dev] [PATCH v3 2/5] dpif-netdev: Trigger parallel pmd reloads.

On 09.07.2019 11:00, David Marchand wrote: > > > On Tue, Jul 9, 2019 at 9:36 AM Ilya Maximets > wrote: > > On 09.07.2019 10:14, David Marchand wrote: > > We can skip the non-PMD thread in this loop for consistency with > dp_netdev_reload_pmd__() if this i

Re: [ovs-dev] [PATCHv14 2/2] netdev-afxdp: add new netdev type for AF_XDP.

On 09.07.2019 12:56, Eelco Chaudron wrote: > Hi William, > > I tried running my PVP test, and with this patchset, I could not see the > deadlock I used to see with all the previous patches. > > However there is still the issue that the maximum packet size supported is > 1028 bytes, all larger p

Re: [ovs-dev] [PATCH v3 2/5] dpif-netdev: Trigger parallel pmd reloads.

> On 09.07.2019 11:00, David Marchand wrote: > > > > > > On Tue, Jul 9, 2019 at 9:36 AM Ilya Maximets > wrote: > > > > On 09.07.2019 10:14, David Marchand wrote: > > > We can skip the non-PMD thread in this loop for consistency with > dp_netdev_reload_pmd__()

Re: [ovs-dev] [PATCHv14 2/2] netdev-afxdp: add new netdev type for AF_XDP.

On 09.07.2019 0:40, William Tu wrote: > On Mon, Jul 8, 2019 at 1:50 PM William Tu wrote: >> >> Hi Ilya, >> >> Thanks for all the feedback! >> >> On Fri, Jul 5, 2019 at 8:32 AM Ilya Maximets wrote: >>> >>> On 01.07.2019 19:08, William Tu wrote: The patch introduces experimental AF_XDP support

Re: [ovs-dev] [PATCHv14 2/2] netdev-afxdp: add new netdev type for AF_XDP.

On 9 Jul 2019, at 12:21, Ilya Maximets wrote: On 09.07.2019 12:56, Eelco Chaudron wrote: Hi William, I tried running my PVP test, and with this patchset, I could not see the deadlock I used to see with all the previous patches. However there is still the issue that the maximum packet size

Re: [ovs-dev] [PATCH] ovsdb-server: drop all connections on read/write status change

Glad to see this being fixed, thanks Daniel. On Tue, Jul 9, 2019 at 11:22 AM Daniel Alvarez wrote: > > Prior to this patch, only db change aware connections were dropped > on a read/write status change. However, current schema in OVN does > not allow clients to monitor whether a particular DB cha

Re: [ovs-dev] [PATCHv14 2/2] netdev-afxdp: add new netdev type for AF_XDP.

On 09.07.2019 13:34, Eelco Chaudron wrote: > > > On 9 Jul 2019, at 12:21, Ilya Maximets wrote: > >> On 09.07.2019 12:56, Eelco Chaudron wrote: >>> Hi William, >>> >>> I tried running my PVP test, and with this patchset, I could not see the >>> deadlock I used to see with all the previous patc

Re: [ovs-dev] [PATCH v3 3/3] ovn: Send GARP for router port IPs of a router port connected to bridged logical switch

On 01.07.2019 10:43, nusid...@redhat.com wrote: > From: Numan Siddique > > This patch handles sending GARPs for > > - router port IPs of a distributed router port > > - router port IPs of a router port which belongs to gateway router >(with the option - redirect-chassis set in Logical_Rou

Re: [ovs-dev] [PATCH RFC v2 4/8] netdev-offload-tc: Implement netdev tc flush via tc filter del

On 7/4/2019 5:28 PM, Paul Blakey wrote: > +static void > +del_ufid_tc_mapping(const ovs_u128 *ufid) > +{ > +ovs_mutex_lock(&ufid_lock); > +del_ufid_tc_mapping_unlocked(ufid); > +ovs_mutex_lock(&ufid_lock); > } > should be ovs_mutex_unlock at end of func. ___

[ovs-dev] IHR ANSRUCH!!

ABOGADOS VIDAL y PALMER Paseo de la Delicias 29 , Madrid 28045 SPAIN TEL: 0034 632 618 927 FAX: 0034 912 916 314 EMAIL: abogadosvidalpal...@consultant.com AKTENZEIHEN: AG/57-076/18-ES KUNDENNUMMER: MD-ES/YGHD-502 Abschließende Mitteilung für die Zahlung des nicht beanspruchten Preisgeldes Wir mö

Re: [ovs-dev] [PATCH v3 3/3] ovn: Send GARP for router port IPs of a router port connected to bridged logical switch

On Tue, Jul 9, 2019 at 4:37 PM Ilya Maximets wrote: > On 01.07.2019 10:43, nusid...@redhat.com wrote: > > From: Numan Siddique > > > > This patch handles sending GARPs for > > > > - router port IPs of a distributed router port > > > > - router port IPs of a router port which belongs to gateway

[ovs-dev] [PATCH v10 0/5] dpcls func ptrs & optimizations

Hey All, Here a v10 of the DPCLS Function Pointer patchset, as has been presented at OVS Conf in Nov '18, and discussed on the ML since then. I'm aware of the soft-freeze for 2.12, I feel this patchset has had enough reviews/versions/testing to be merged in 2.12. Thanks Ilya and Ian for review c

[ovs-dev] [PATCH v10 1/5] dpif-netdev: implement function pointers/subtable

This allows plugging-in of different subtable hash-lookup-verify routines, and allows special casing of those functions based on known context (eg: # of bits set) of the specific subtable. Signed-off-by: Harry van Haaren Tested-by: Malvika Gupta --- v10: - Fix capitalization of comments, and p

[ovs-dev] [PATCH v10 2/5] dpif-netdev: move dpcls lookup structures to .h

This commit moves some data-structures to be available in the dpif-netdev.h header. This allows specific implementations of the subtable lookup function to include just that header file, and not require that the code exists in dpif-netdev.c Signed-off-by: Harry van Haaren Tested-by: Malvika Gupta

[ovs-dev] [PATCH v10 3/5] dpif-netdev: split out generic lookup function

This commit splits the generic hash-lookup-verify function to its own file. In doing so, we must move some MACRO definitions to dpif-netdev.h Signed-off-by: Harry van Haaren Tested-by: Malvika Gupta --- v10: - Rebase fixups from previous patch changes - Add copyright as suggested in review (Ia

[ovs-dev] [PATCH v10 4/5] dpif-netdev: refactor generic implementation

This commit refactors the generic implementation. The goal of this refactor is to simply the code to enable "specialization" of the functions at compile time. Given compile-time optimizations, the compiler is able to unroll loops, and create optimized code sequences due to compile time knowledge o

[ovs-dev] [PATCH v10 5/5] dpif-netdev: add specialized generic scalar functions

This commit adds a number of specialized functions, that handle common miniflow fingerprints. This enables compiler optimization, resulting in higher performance. Below a quick description of how this optimization actually works; "Specialized functions" are "instances" of the generic implementatio

[ovs-dev] [PATCH v2 0/3] OVN: Add IGMP support

This series introduces support for IGMP Snooping and IGMP Querier. IGMP versions v1-v3 are supported for snooping and IGMP queries originated by ovn-controller are general IGMPv3 queries. The rationale behind using v3 for querier is that it's backward compatible with v1-v2. The majority of the cod

[ovs-dev] [PATCH v2 1/3] packets: Add IGMPv3 query packet definitions

Signed-off-by: Dumitru Ceara --- lib/packets.c | 44 lib/packets.h | 19 ++- 2 files changed, 62 insertions(+), 1 deletion(-) diff --git a/lib/packets.c b/lib/packets.c index a8fd61f..ab0b1a3 100644 --- a/lib/packets.c +++ b/lib/pa

[ovs-dev] [PATCH v2 2/3] OVN: Add IGMP SB definitions and ovn-controller support

A new IP_Multicast table is added to Southbound DB. This table stores the multicast related configuration for each datapath. Each row will be populated by ovn-northd and will control: - if IGMP Snooping is enabled or not, the snooping table size and multicast group idle timeout. - if IGMP Querier

[ovs-dev] [PATCH v2 3/3] OVN: Add ovn-northd IGMP support

New IP Multicast Snooping Options are added to the Northbound DB Logical_Switch:other_config column. These allow enabling IGMP snooping and querier on the logical switch and get translated by ovn-northd to rows in the IP_Multicast Southbound DB table. ovn-northd monitors for changes done by ovn-co

Re: [ovs-dev] [PATCH v3] ovn-northd: Add the option to pause and resume

Acked-by: Mark Michelson On 7/9/19 1:10 AM, nusid...@redhat.com wrote: From: Numan Siddique This patch adds 3 unixctl socket comments - pause, resume and is-paused. Usage: ovs-appctl -t ovn-northd pause/resume/is-paused This feature will be useful if the CMS wants to - deploy OVN DB serv

Re: [ovs-dev] [PATCH v10 3/5] dpif-netdev: split out generic lookup function

Bleep bloop. Greetings Harry van Haaren, I am a robot and I have tried out your patch. Thanks for your contribution. I encountered some error that I wasn't expecting. See the details below. checkpatch: ERROR: Improper whitespace around control block #278 FILE: lib/dpif-netdev.h:105: #define N

[ovs-dev] Soft freeze exception request for OVN IGMP support in 2.12

Hi, Would it be possible to have an exception for IGMP support in 2.12 once review is done? Ben already did a preliminary review of v1 and I posted v2 [1] today. Thanks, Dumitru [1] https://patchwork.ozlabs.org/project/openvswitch/list/?series=118544 _

[ovs-dev] [PATCH 1/2] compat: ip6_gre: fix possible use-after-free in ip6erspan_rcv

Upstream commit: commit 2a3cabae4536edbcb21d344e7aa8be7a584d2afb Author: Lorenzo Bianconi Date: Sat Apr 6 17:16:53 2019 +0200 net: ip6_gre: fix possible use-after-free in ip6erspan_rcv erspan_v6 tunnels run __iptunnel_pull_header on received skbs to remove erspan header

[ovs-dev] [PATCH 2/2] datapath: fix csum updates for MPLS actions

Upstream commit: commit 0e3183cd2a64843a95b62f8bd4a83605a4cf0615 Author: John Hurley Date: Thu Jun 27 14:37:30 2019 +0100 net: openvswitch: fix csum updates for MPLS actions Skbs may have their checksum value populated by HW. If this is a checksum calculated over the en

[ovs-dev] Kreditangebot

Sehr geehrte Damen und Herren, Benötigen Sie finanzielle Hilfe? Sind Sie es auch leid, mit Banken oder Finanzdienstleistern um einen Kredit zu suchen? Um Sicherheiten zu erhalten, um das Darlehen zu 150% zu sichern? Brauchen Sie einen Kredit, um eine Investition zu tätigen, kurzfristige Schul

Re: [ovs-dev] [patch v5] conntrack: Optimize recirculations.

Thanks for the review ! On Mon, Jul 8, 2019 at 2:58 PM Ben Pfaff wrote: > On Sun, Jun 09, 2019 at 07:32:03AM -0700, Darrell Ball wrote: > > Cache the 'conn' context and use it when it is valid. The cached 'conn' > > context will get reset if it is not expected to be valid; a negative test > > i

[ovs-dev] [PATCH v4 0/5] Quicker pmd threads reloads

We have been testing the rebalance code in different situations while having traffic going through OVS. Those tests have shown that part of the observed packets losses is due to some time wasted in signaling/waiting for the pmd threads to reload their polling configurations. This series is an atte

[ovs-dev] [PATCH v4 1/5] dpif-netdev: Convert exit latch to flag.

No need for a latch here since we don't have to wait. A simple boolean flag is enough. The memory order on the reload flag is changed to rel-acq ordering to serve as a synchronisation point between the pmd threads and the control thread that asks for termination. Fixes: e4cfed38b159 ("dpif-netdev

[ovs-dev] [PATCH v4 2/5] dpif-netdev: Trigger parallel pmd reloads.

pmd reloads are currently serialised in each steps calling reload_affected_pmds. Any pmd processing packets, waiting on a mutex etc... will make other pmd threads wait for a delay that can be undeterministic when syscalls adds up. Switch to a little busy loop on the control thread using the existi

[ovs-dev] [PATCH v4 4/5] dpif-netdev: Only reload static tx qid when needed.

pmd->static_tx_qid is allocated under a mutex by the different pmd threads. Unconditionally reallocating it will make those pmd threads sleep when contention occurs. During "normal" reloads like for rebalancing queues between pmd threads, this can make pmd threads waste time on this. Reallocating t

[ovs-dev] [PATCH v4 5/5] dpif-netdev: Catch reloads faster.

Looking at the reload flag only every 1024 loops can be a long time under load, since we might be handling 32 packets per rxq, per iteration, which means up to poll_cnt * 32 * 1024 packets. Look at the flag every loop, no major performance impact seen. Signed-off-by: David Marchand Acked-by: Eelc

[ovs-dev] [PATCH v4 3/5] dpif-netdev: Do not sleep when swapping queues.

When swapping queues from a pmd thread to another (q0 polled by pmd0/q1 polled by pmd1 -> q1 polled by pmd0/q0 polled by pmd1), the current "Step 5" puts both pmds to sleep waiting for the control thread to wake them up later. Prefer to make them spin in such a case to avoid sleeping an undetermin

[ovs-dev] [PATCHv15 1/2] ovs-thread: Add pthread spin lock support.

The patch adds the basic spin lock functions: ovs_spin_{lock, try_lock, unlock, init, destroy}. OSX does not support pthread spin lock, so make it linux only. Signed-off-by: William Tu --- include/openvswitch/thread.h | 22 ++ lib/ovs-thread.c | 31 +++

[ovs-dev] [PATCHv15 2/2] netdev-afxdp: add new netdev type for AF_XDP.

The patch introduces experimental AF_XDP support for OVS netdev. AF_XDP, the Address Family of the eXpress Data Path, is a new Linux socket type built upon the eBPF and XDP technology. It is aims to have comparable performance to DPDK but cooperate better with existing kernel's networking stack.

Re: [ovs-dev] [PATCH 1/2] compat: ip6_gre: fix possible use-after-free in ip6erspan_rcv

On Tue, Jul 9, 2019 at 8:25 AM Greg Rose wrote: > > Upstream commit: > commit 2a3cabae4536edbcb21d344e7aa8be7a584d2afb > Author: Lorenzo Bianconi > Date: Sat Apr 6 17:16:53 2019 +0200 > > net: ip6_gre: fix possible use-after-free in ip6erspan_rcv > > erspan_v6 tunnels run __

Re: [ovs-dev] [PATCH 2/2] datapath: fix csum updates for MPLS actions

On Tue, Jul 9, 2019 at 8:25 AM Greg Rose wrote: > > Upstream commit: > commit 0e3183cd2a64843a95b62f8bd4a83605a4cf0615 > Author: John Hurley > Date: Thu Jun 27 14:37:30 2019 +0100 > > net: openvswitch: fix csum updates for MPLS actions > > Skbs may have their checksum value

[ovs-dev] Soft Freeze exception request for OVN controller events support in 2.12

I'd like to request a soft freeze exception for OVN controller events [1] as submitted by Lorenzo. Thank you, Mark Michelson [1] http://patchwork.ozlabs.org/project/openvswitch/list/?series=113941 ___ dev mailing list d...@openvswitch.org https://mail

Re: [ovs-dev] [PATCH] tnl-neigh-cache: Purge learnt neighbors when port/bridge is deleted

My comments inline: On Mon, Jul 8, 2019 at 9:41 AM Flavio Leitner wrote: > On Tue, Jul 02, 2019 at 08:50:16PM -0400, Vasu Dasari wrote: > > Hi Flavio, > > > > I am trying to emulate the test case scenario you mentioned earlier, > where > > in we need to clean you neighbor cache when an external

[ovs-dev] Why is ovs DPDK much worse than ovs in my test case?

Hi, all I just use ovs as a static router in my test case, ovs is ran in vagrant VM, ethernet interfaces uses virtio driver, I create two ovs bridges, each one adds one ethernet interface, two bridges are connected by patch port, only default openflow rule is there. table=0, priority=0 acti

[ovs-dev] [PATCH 1/2] ovn-performance.at: Missing steps for connecting LS to LR.

From: Han Zhou The test creates 2 logical switches and connect them with a logical router. However, it didn't set the option "router-port", so the 2 LS datapaths were not connected. This results in missing test coverage for port-binding incremental processing: assume I-P has a bug and port-bindin

[ovs-dev] [PATCH 2/2] ovn-performance.at: Fix syntax error in ACL.

From: Han Zhou This doesn't impact the effectiveness of the test but just fix an obvious error in ACL syntax which was noticed when looking at test logs. Signed-off-by: Han Zhou --- tests/ovn-performance.at | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tests/ovn-perfo

[ovs-dev] [PATCH] flow: Wildcard UDP ports when using SYMMETRIC_L4 hash for select groups.

UDP source and destination ports are not used to derive the hash index used for selecting the bucket in case of SYMMETRIC_L4 hash based select groups. However, they are un-wildcarded in the megaflow entry match criteria. This results in distinct megaflow entry being created for each pair of UDP sou